Surg Endosc (2006) 20: 191–198 DOI: 10.1007/s00464-005-0436-7 Ó Springer Science+Business Media, Inc. 2005
Analysis of the SAGES Outcomes Initiative groin hernia database V. Velanovich,1 P. Shadduck,2 L. Khaitan,3 J. Morton,4 G. Maupin,5 L. W. Traverso6 for the SAGES Outcomes Committee 1 2 3 4 5 6
Division of General Surgery, K-8, Henry Ford Hospital, 2799 West Grand Boulevard, Detroit, MI 48202, USA Durham Surgical Associates, Durham, NC, USA Emory University, Atlanta, GA, USA Stanford University, Palo Alto, CA, USA Whisper Communications, Kingston, WA, USA Virginia-Mason Medical Center, Seattle, WA, USA
Received: 17 June 2005/Accepted: 12 September 2005/Online publication: 6 December 2005
Abstract Background: In 1999, the Society of American Gastrointestinal Endoscopic Surgeons (SAGES) introduced the SAGES Outcomes Initiative as a way for its members to track their own outcomes. It contains perioperative and postoperative data on nearly 20,000 operations. This report provides a descriptive analysis of the groin hernia database. Methods: The SAGES Outcomes Initiative database was accessed for all groin hernia cases from September 1999 to February 2005. The data from the preoperative, intraoperative, and postoperative entries were summarized. These data are purely descriptive and no statistical analysis was done. Results: The hernia registry contains 1,607 entries, with 1,070 follow-up entries. Males comprised 85% of patients, 63% were employed, 62% had at least one comorbidity, with 84% ASA class I or II. Primary, unilateral hernia accounted for 86% of cases, whereas 14% were recurrent, 11% bilateral, 6% incarcerated, and 3% required emergency repair. The operating surgeon was the attending surgeon in 83% of cases. Anesthetic techniques were general anesthesia in 74% of cases, regional in 7%, and local in 34%, with only 16% of cases local only. Most patients had symptomatic hernias and symptoms were improved in more than 95% of patients. Most repairs were open, although 45% were endoscopic. The most frequently cited postoperative event was significant bruising (6%), with more than 99% of complications being class I or II. More than 95% of patients were able to return to work by the first postoperative visit. Patients who underwent endoscopic repair were
Presented at the Society of American Gastrointestinal Endoscopic Surgeons (SAGES) meeting, Hollywood, FL, USA, 14 April 2005 Correspondence to: V. Velanovich
reported to have fewer days of narcotic use than patients undergoing open repairs (0 vs 3). Conclusions: First analysis of the SAGES Outcomes Initiative groin hernia database demonstrates that (a) this is one of the largest prospective; voluntary hernia registries; (b) missing data are infrequent; and (c) the data are similar to published data from national, mandatory registries and randomized trials. Although the SAGES Outcomes Initiative is a voluntary registry, initially designed for surgeon self-assessment, and it therefore has the potential for methodological concerns inherent to voluntary registries, the findings from this first analysis are encouraging. Efforts are ongoing to simplify data entry (PDA), refine data parameters, increase surgeon participation, and determine the role of data audit and thereby the potential for clinical research. Key words: Hernia — Surgical< Technical — Technical < Endoscopy — Complications — Clinical papers / trials /research
Surgeons have long appreciated the importance of evaluating the safety, efficacy, and value of operative techniques and management strategies. Governmental agencies, payors, and others have begun to appreciate the importance of these efforts as well, resulting in a greater emphasis on outcomes research and evidence-based decision making. The recent publication by Medicare of institutional outcomes on the Internet and the early efforts at implementing ‘‘pay for performance’’ strategies for physicians have carried this movement to new levels. Two decades ago, the Society of Thoracic Surgeons took a bold step forward by developing, implementing, and promoting a voluntary, prospective, national registry of outcomes for cardiac surgery, but now including other thoracic procedures as well [19]. This particular
192
Fig. 1. The Initial Input Surgical log.
database has been refined, validated, and matured to the point that it now influences health care decision making and, for Blue Cross/Blue Shield, physician contracting and reimbursement [19]. In 1999, the Society of American Gastrointestinal Endoscopic Surgeons (SAGES) launched the Outcomes Initiative to provide its members a vehicle to collect and track their outcomes. The Outcomes Initiative is the first voluntary, prospective, Health Insurance Portability and Accountability Act (HIPAA)-compliant registry open to gastrointestinal/laparoscopic surgeons. The purposes for establishing this registry were (a) to provide surgeons a mechanism that facilitates real-time selfassessment and self-improvement; (b) to provide individual surgeons a mechanism to record their own outcomes (with the potential to utilize this information for credentialing and contracting, when necessary); and (c) to provide the gastrointestinal/laparoscopic surgical family an opportunity to help determine and establish national benchmarks, information necessary for an informed dialogue with governmental agencies, payors, and other interested parties. Previous publications have introduced the database to the membership [11] and have compared its data for antireflux surgery to the National Inpatient Sample [15]. However, neither of these publications have truly reported the breadth of the data that the initiative records. The purpose of this report is to assess the practice patterns of the SAGES members who input cases into the database for hernia surgery. By this assessment, our goal is to improve the
collection of data and determine if the results are broadly comparable to other published data.
Materials and methods The SAGES Outcomes Initiative was established in May 1999 through the SAGES Outcomes Task Force in an effort to provide an outcomes reporting system for its members. Any SAGES member is eligible to register and input case information. Although participation is voluntary, once a member is enrolled, it is expected that all cases will be recorded in a consecutive manner. Case information is entered through the Outcomes link through the SAGES web site. The web site is managed by Whisper Communications (Kingston, WA, USA). The service is free to any SAGES member. After obtaining a password and stating adherence to HIPAA rules, enrollees may access the web site at any time. Preoperative and operative data common to all operations are entered using the Initial Surgical Log form (Fig. 1) and those specific to hernias in the Initial Hernia Log form (Fig. 2). On the first postoperative visit, the Surgical Follow-Up form (Fig. 3) and Hernia Follow-Up form (Fig. 4) are completed. Each data form was constructed by a panel of experts, beta tested, and then implemented in the database. Complication severity was based on the classification scheme of Clavien et al. [4]. The SAGES hernia registry was accessed for all inguinal hernia cases from September 1999 to February 2005. Descriptive statistics for each data point were calculated, including missing or nonsensical data.
Results A total of 1,607 initial cases, with 1,070 follow-up cases, were identified in the hernia database. The mean
193
Fig. 2. The Initial Input Hernia log.
Fig. 3. The Surgical Follow-Up form.
age (±SD) was 54 ± 16 years. The group was 85% male (1,353 of 1,585), with 22 missing data points (1.4%). Table 1 presents preoperative work status, ASA classification, and preoperative comorbidities. Emergency surgery was reported in 2.8% of cases (46 of 1,589, with 18 missing data points). Anesthesia
techniques used included general in 73.5% (1,181), spinal/epidural in 6.8% (110), general and spinal in 0.8% (13), and local in 33.5% (538), with local only in 16%. Only five cases had no anesthesia entered. Average estimated blood loss was 13 ± 22 ml (data skewed to the left), and mean operating time was 58 ± 37 min.
194
Fig. 4. The Hernia Follow-Up form.
Table 1. Preoperative patients dataa Work status
%
ASA classification
%
Comorbidities
%
Not specified Office Physical labour Student Unemployed Retired Disabled Professional Other
42 22.1 24.8 3.5 1.9 26.8 2.7 7.5 4.7
I II III IV V
40.3 44.3 14.6 0.8 0
None Cardiac Pulmonary Kidney Liver Alcohol Diabetes Hypertension Malignancy CNS Obesity
37.8 15.1 14.9 1.6 1.9 1.6 4.9 24.4 3.4 4.2 8.7
a There were 28 missing entities for work status and 72 missing entries for ASA classification. For work status and comorbidities, more that one could be selected
Fig. 5. Distribution of ICD-9 diagnostic codes: 550.9, inguinal hernia, without mention of obstruction or gangrene; 550.1, inguinal hernia, with obstruction, without mention of gangrene; 550.91, unilateral, recurrent; 550.92, bilateral, not recurrent; 550.93, bilateral, recurrent; 553.1, umbilical hernia; 553.21, incisional hernia; 552, other hernia of abdominal cavity.
A robot was used in 141 of 1,247 cases (11%). Figure 5 presents the distribution of diagnostic codes reported. Figure 6 presents the distribution of procedure codes reported. Table 2 reports the experience of operating
and assistant surgeons. Previous hernia operations were reported in 14.4% of patients (222 of 1,543). Preoperative symptomatic hernias were reported in 96.5% of patients (1,500 of 1,555).
195
Fig. 6. Distribution of CPT codes: 49505, repair initial inguinal hernia, age >5 years, reducible; 49507, incarcerated or strangulated; 49520, repair recurrent inguinal hernia; 49525, repair sliding inguinal hernia; 49560, repair incisional hernia; 49561, repair incarcerated incisional hernia; 49650, laparoscopic repair initial inguinal hernia; 4965050, laparoscopic bilateral repair; 49561, laparoscopic repair recurrent inguinal hernia; 49553, repair incarcerated femoral hernia. Table 2. Operating and assistant surgeonsa
Attending Fellow R4–5 R1–3 Nonphysician
Table 3. Use of disposable instruments
Operating surgeon (%)
Assistant surgeon (%)
Instrument
No. used
83.4 2.8 6.6 7.1 —
28.1 1.8 10.9 8.5 50.7
None Trocars Stapler Balloon dissector Tacker Other
500 657 275 486 653 6
a Missing data for operating surgeons, 84; missing data for assistant surgeon, 460
Table 4. Procedures performed and operating findings Procedure performeda
%
Type of mesh
%
Type of hernia
%
Open mesh Open patch Open plug Open suture only TEP TAPP Lap Patch Neither
37.1 2.5 12.8 5.2 33.6 9.8% 3.1% 1.6%
Polypropylene PTFE Composite Other None
85.6 1.1 5.0 3.5 4.7
Indirect Direct Femoral Bilateral Nonreducile
48.5 34.2 2.3 9.9 5.1
a
n > 1607 because some entries had more than one procedure performed
Table 5. Postoperative events Postoperative event
Significant brusing
Table 6. Postoperative symptomatic change and work activity
Frequency Complication (%) severity 8.9
Hematoma requiring 0.8 drainage Persistent pain 3.6 Wound infection 0.6 Persistent numbness 2.0 Urinary 1.7 Deep venous thrombosis 0 Other complications 12.0
Frequency (%)
N/A 41.4 (no complication) I 3.0 II III IV None of the above
0.2 0.5 0.2 52.0
Table 3 lists the number of disposable instruments used. The mean number of disposable trocars used was three. Table 4 lists the operations and operative findings. The sum of the procedures performed (1,686) is larger than the total number of cases in the database
Symptomatic change
%
Work status
%
Marked improvement Somewhat improved No change Somewhat worse Much worse
77.6 18.8 2.0 1.2 0.4
Full activity Partial activity Unable due to procedure Unable due to other reasons
64.5 31.0 2.9 1.6
because more than one type of repair was chosen for many entries. However, 55% of entries (927 of 1,686) were coded for open repairs, whereas 45% were coded for endoscopic repairs. Of the operations done endoscopically, 63 entries (8.3%) reported converting to the open approach. Surgeons reported no intraoperative events in 1,403 entries (87.3%), bleeding that slowed the case in 17 (1.1%), and pneumoperitoneum in 44 (2.7%). Average follow-up was 11 ± 4 days. Table 5 lists the preoperative events and complication severity. When
196 Table 7. Postoperative event occurrence rates based on repair type Occurrrence (%) Repair
Bruising
Hematoma
Infection
Pain
Numbness
DVT
Urinary
Other
Open Lap
10.3 6.3
0.3 0.8
0 0.3
4.6 1.9
1.9 0.8
0 0
0.3 2.4
11.5 11.7
Table 8. Number of postoperative oral narcotic days based on procedure type No. of days Procedure
Mean (SD)
Median
Range
IQ Range
Open (n = 301) Lap (n = 255) Both (n = 41)
3.5 (3.0) 1.1 (2.4) 1.4 (5.5)
3.0 0 0
0–25 0–21 0–30
2–5 0–2 0–0
Table 9. Comparison of SAGES Outcomes Initiative hernia registry to other databases Information
SAGES
Scotland [21]
UK [19]
Sweden [16, 17]
US Army [9]
Type of data No. of patients Gender Age Employment ASA class Hernia type Anesthesia type Repair type Operation time Operating surgeon Narcotic use Return to activity Complications Severity Patient follow-up Hernia recurrence
Registry 1,607 Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes No
Audit 3,150 Yes Yes No Yes Yes Yes Yes Yes Yes No No No No Yes No
Audit 5,038 No No No Yes Yes Yes Yes No Yes No No No No Yes No
Registry 24,310 Yes No No No Yes Yes Yes Yes No No No No No No No
Database 36,250 Yes Yes No No Yes No No No No No No Yes No No No
complications occurred, most were grade I. Readmissions due to the procedure were recorded in 16 entries, whereas 849 did not report readmission (205 data points missing; 19.2%). Table 6 reports the postoperative symptomatic change and postoperative work status. Patients reported first leaving home at a mean of 2 ± 2 days, and patients used narcotics for a mean of 2 ± 3 days. These data were skewed to the left. Tables 7 and 8 compare the rates of postoperative event occurrence and postoperative work activity by open and endoscopic repairs.
Discussion Insights that can be gleaned from review of these data can be segregated into three categories. First, the quality of the database can be assessed. Second, the practice patterns of the SAGES members entering cases can be elucidated. Third, some broad comparisons can be made. Lastly, some decisions regarding the future
direction for and refinement of the database can be determined. The quality of the data appears to be good, particularly for the relatively early stage of its development. All the items chosen for measurement were done so by SAGES surgeons expert in issues involved in hernia surgery. Therefore, these are the patient characteristics, procedure details, and outcomes that matter most. As such, SAGES members would have an interest in accurately recording these items. In fact, given the number of items that surgeons are asked to record, there are relatively few missing data points. The frequency of missing data points is in the few percent range. In addition, when SAGES members register to input data into the database, they agree to participate in an audit. Nevertheless, it must be stated that audits of individual surgeons to date have not been done. Therefore, we have no independent confirmation of the truthfulness of the data. On the other hand, because the data are recorded by the surgeon, rather than administrative personnel, we can assume what is recorded is accurate. The operating surgeon knows the type of hernia, type of repair, mesh,
197
etc., which a coder may not believe is important from the standpoint of billing but is very important from the standpoint of outcome. When the Outcomes Initiative database was compared to the National Inpatient Sample for antireflux surgery, there were very similar complication rates, for example [15]. In fact, for the hernia registry, the types of operations and outcomes are very similar to published reports, again implying truthfulness in recording outcomes by SAGES members. The practice patterns of the SAGES members contributing to the database were somewhat surprising. First, despite SAGES being an endoscopic/laparoscopic organization, most groin hernias were done using the open technique (55%). Even so, the majority of cases were done with a general anesthetic (74.3%). It is difficult to determine the frequency of local-only anesthesia because combinations of local and general and/or regional were used frequently. The vast majority were first-time unilateral hernias (Fig. 4), with more than 95% being symptomatic. Postoperatively, significant bruising and persistent pain were the most common adverse events (Table 5). Most patients were able to return to full activity/work (Table 6). The attending surgeon was the operating and assistant surgeon in most cases, implying that most surgeons entering data worked without residents (and, therefore, in nonteaching hospitals). These data are comparable to other published reports, again acting as a surrogate quality control measure [1, 2, 5, 8, 10, 12, 14]. Some broad comparisons can be made between the type of operation. The comparison of most interest is that of the open versus the endoscopic approach. Table 7 demonstrates that the frequency of adverse postoperative events was similar between both groups. The open approach may have a slightly higher incidence of bruising, persistent pain, and persistent numbness, whereas the endoscopic approach had a slightly higher incidence of urinary complications. Table 8 demonstrates that patients who underwent the endoscopic approach had fewer days of postoperative narcotic use compared to those who underwent the open approach. Comparisons of subgroups within the SAGES database need to be performed with caution due to the methodologic issues previously acknowledged. Nevertheless, the data for open mesh and laparoscopic repairs are consistent with data from randomized, controlled trials [3, 6, 7, 13, 20, 22]. Although not a direct purpose of this study, it would be appropriate to broadly compare the SAGES hernia database to others. Comparisons of the data from the SAGES hernia database were made to published data from national registries (England, Scotland, Sweden, and the U.S. Army) [9, 16–18] and to randomized controlled trials (Table 9). The first author had the opportunity to review the U.S. Army Patient Administrations Systems and Biostatistics Activity database for abdominal hernia repairs in 1990 [9]. During a 5-year period, 36,250 hernia operations were done in U.S. Army hospitals. However, the breath of data is not great because hernias were generally categorized by site and the presence or absence of obstruction or gangrene, and the only outcome was mortality. Compared to the SAGES
database, there is a larger sample size but not much clinical detail. The Medicare database that is frequently used only records data for Medicare patients, who are generally older. Therefore, this database may not apply to younger patients. The SAGES database covers all patients treated by SAGES surgeons participating in the initiative; therefore, although age may not be a consideration, other bias may be because SAGES surgeons participating may not be reflective of the larger population of surgeons. The National Inpatient Sample only records inpatient data; this clearly is not appropriate for hernia repair because most patients are operated on in the outpatient setting. Therefore, all databases have limitations, and care in interpreting the data is needed. Finally, current priorities for the Outcomes Initiative in general, and the hernia database specifically, include the following: (a) simplification of data entry (a revised PDA software format was recently released); (b) refinement of data parameters; (c) expanded participation by member surgeons; and (d) determination of the desirability of, methodology for, and time frame for auditing data, thereby determining the feasibility of utilizing the initiative as a resource for clinical research. In conclusion, analysis of the SAGES Outcomes Initiative hernia database reveals the following: (a) This is one of the largest prospective hernia registries, (b) missing data points are relatively few, and (c) the data are similar to those of published national registries and randomized clinical trials. Although the SAGES Outcomes Initiative is a voluntary registry (initially designed for surgeon self-assessment) and it thus suffers from the methodological concerns inherent to voluntary registries, the findings from this first analysis are encouraging. Efforts are ongoing to simplify data entry, refine data parameters, increase surgeon participation, and determine the role of data audit and thereby the potential for clinical research. References 1. Atkinson HD, Nicol SG, Purkayastha S, Paterson-Brown S (2004) Surgical management of inguinal hernia: retrospective cohort study in southeastern Scotland, 1985–2001. Br Med J 329: 1315– 1316 2. Aufenacker TJ, de Lange DH, Burg MK, Kuiken BW, Hensen EF, Schoo IG, Gouma DJ, Simons MP (2005) Hernia surgery changes in the Amsterdam region 1994–2001: decrease in operations for recurrent hernia. Hernia 9: 46–50 3. Bozuk M, Schuster R, Stewart D, Hicks K, Greaney G, Waxman K (2003) Disability and chronic pain after open mesh and laparoscopic inguinal hernia repair. Am Surg 69: 839–841 4. Clavien PA, Sanabria JR, Strasberg SM (1992) Proposed classification of complications of surgery with examples of utility in cholecystectomy. Surgery 111: 518–526 5. de Lange DH, Aufenacker TJ, Roest M, Simmermacher RK, Gouma DJ, Simons MP (2005) Inguinal hernia surgery in the Netherlands: a baseline study before the introduction of the Dutch guidelines. Hernia 9: 172–177 6. Douek M, Smith G, Oshowo A, Stoker DL, Wellwood JM (2003) Prospective randomized controlled trial of laparoscopic versus open inguinal hernia mesh repair: five year follow up. Br Med J 326: 1012–1013 7. Fujita F, Lahmann B, Otsuka K, Lyass S, Hiatt JR, Phillips EH (2004) Quantification of pain and satisfaction following laparoscopic and open hernia repair. Arch Surg 139: 596–600
198 8. Grant AM, Scott NW, OÕDwyer PJ, MRC laparoscopic Groin Hernia Trial Group (2004) Five-year follow-up of a randomized trial to assess pain and numbness after laparoscopic or open repair of groin hernia. Br J Surg 91: 1570–1574 9. Heydorn WH, Velanovich V (1990) A five-year U.S. Army experience with 36,250 abdominal hernia repairs. Am Surg 56: 596–600 10. Kehlet H, Bay Nielsen M (2005) Anaesthetic practice for groin hernia repair—a nation-wide study in Denmark 1998–2003. Acta Anaesthesiol Scand 49: 143–146 11. Khaitan L, Apelgren K, Hunter J, Traverso LW (2003) A report on the Society of American Gastrointestinal Endoscopic Surgeons Outcomes Initiative: what have we learned and what is the potential? Surg Endosc 17: 365–370 12. Kumar S, Foo Wong P, Melling A, Leaper DJ (2004) Surgical site infection after groin hernia repair. Br J Surg 91: 105–111 13. McCormack K, Scott NW, Go PM, Ross S, Grant AM, EU Hernia Trialists Collaboration (2003) Laparoscopic techniques versus open techniques for inguinal hernia repair. Cochrane Database Syst Rev, CD001785 14. Mikkelsen T, Werner MU, Lassen B, Kehlet H (2004) Pain and sensory dysfunction 6 to 12 months after inguinal herniotomy. Anesth Analg 99: 146–151
15. Morton JM, Galanko JA, Soper NJ, Low DE, Hunter J, Traverso LW (2005) NIS vs SAGES: a comparison of national and voluntary databases. Surg Endosc, in press 16. Nilsson E, Haapaniemi S (1998) Hernia registries and specialization. Surg Clin North Am 78: 1141–1155 17. Nilsson E, Haapaniemi S (2002) Assessing the quality of hernia repair. In: Nyhus L, Condon RE (eds) Hernia, 5th ed. 18. OÕRiordan DC, Kingsnorth AN (1998) Audit of patient outcomes after herniorrhaphy. Surg Clin North Am 78: 1129–1139 19. Orringer MB (2001) STS database activities and you: ‘‘whatÕs in it for me?’’ Ann Thorac Surg 72: 1–2 20. Stengel D, Bauwens K, Ekkernkamp A (2004) Recurrence risks in randomized trials of laparoscopic versus open inguinal hernia repairs: to pool or not to pool (this is not the question). Langenbecks Arch Surg 389: 492–498 21. Tayor EW, Duffy K, Lee K, Hill R, Noone A, Macintyre I, King PM, OÕDwyer PJ (2004) Surgical site infection after groin hernia repair. Br J Surg 19: 105–111 22. Wake B, McCormack K, Fraser C, Vale L, Perez J, Grant A (2005) Transabdominal pre-peritoneal (TAPP) vs totally extraperitoneal (TEP) laparoscopic techniques for inguinal hernia repair. Cochrane Database Syst Rev, CD004703